Maximizing a ULA certified count is a project, not a last minute count, and it works best starting about twelve months before the certification date. Five workstreams run in parallel: discovery, contract review, planned deployment, evidence assembly, and certification preparation, each feeding the defensible number you declare at exit.
By the Meridian advisory team, former Oracle LMS and GLAS licensing analysts. Updated 4 June 2026.
About twelve months before the certification date. The reason is mechanical. Capacity has to be genuinely deployed and running within the term to count, so any deployment you want in the number must be built before the window closes. Discovery across a large estate takes weeks, contract review takes careful reading, and the evidence file is assembled as you go rather than in a panic at the end. A team that starts a year out can run each workstream properly. A team that starts in the final quarter is forced to certify whatever happens to be visible, which is how counts come in low. The economics that justify the effort are set out in why a higher certified count is free value.
Five workstreams, one outcome. Discovery finds the estate, contract review decides what qualifies, deployment brings planned capacity inside the window, evidence proves every number, and certification preparation turns it into a clean declaration.
Everything starts with knowing the true estate. Independent discovery across databases, options, packs, and middleware finds the deployments internal records miss: servers nobody scanned, environments owned outside the project, virtualization that internal inventories understate. The goal is a complete picture, because a count can only be as large as the estate you can see. The tools and methods for this sit in discovery tooling for a ULA count.
Discovery tells you what exists. The contract tells you what qualifies. Cloud counting is contract specific, with many agreements requiring deployments in AWS or Azure to run for a defined period, often 365 continuous days, to count toward the baseline, and some excluding public cloud entirely. The partitioning stance governs how virtualized estates are measured. Customer definition, entity lists, and territory clauses decide which deployments are inside scope. Reading these clauses early shapes the rest of the plan, because they determine where deployment effort pays off.
This is where the count grows. Committed projects are brought forward, disaster recovery is built out, and genuine capacity headroom is provisioned, all inside the window while deployment is still unlimited and free to certify. The discipline is genuineness: every environment must be real, in use or ready for use, and built for a reason you can articulate. We cover how to do this without crossing the line in pre building the next three years of growth.
A number is only worth what you can defend. As each deployment is found or built, the evidence behind it is captured: server lists, discovery output, architecture diagrams, project records, and documented methodology. This file is what protects the count in the years after the exit, when audit risk is highest. Assembling it during the project, rather than reconstructing it later, is the difference between a defensible count and an exposed one.
The final workstream turns the count into a clean declaration. The number is reconciled to the contract, the certification letter is prepared, and the supporting position is readied for the exchange with Oracle. Because a C level executive typically signs the letter, the internal sign off path is mapped early so the close is not held up at the end.
The workstreams overlap rather than run in sequence. The following is an indicative shape for a twelve month runway; your own timeline depends on the size of the estate and the certification date in your contract.
| Months before exit | Primary focus | Outcome |
|---|---|---|
| 12 to 9 | Discovery and contract review | True estate and qualifying rules known |
| 9 to 4 | Planned deployment | Capacity built inside the window |
| 6 to 2 | Evidence assembly | Defensible file behind every number |
| 3 to 0 | Certification preparation | Letter, position, and sign off ready |
The overlap is deliberate. Evidence assembly begins while deployment is still under way, and certification preparation starts before deployment finishes, so nothing is left to the final fortnight.
Take an indicative case. An organisation with a database ULA begins the project twelve months out. Discovery raises the visible estate from a partial inventory to a complete one, adding processors that were simply never scanned. Contract review confirms that a qualifying cloud estate counts and that a virtualized cluster can be measured in the organisation's favour with the right isolation. A deployment programme brings forward a consolidation project and builds out disaster recovery. The count moves from an initial estimate near 1,000 processors to a defensible figure around 1,800, each line evidenced. Support is unchanged throughout. The figures are indicative, but the multiple, well above the first estimate, is the typical reward of running this as a planned project rather than a final count.
A maximization project is the practical expression of a simple truth: the certification window is the cheapest capacity you will ever license, so it deserves a plan. Run the five workstreams in parallel, start early enough to do each properly, and avoid the traps that shrink the reward, which we cover in the maximization mistakes that backfire. The full method, with the rest of the cluster, lives in our pillar, the ULA deployment maximization guide.
Maximization is won in the planning, not the final count. Start about twelve months out, run discovery, contract review, planned deployment, evidence assembly, and certification preparation in parallel, and treat the window as the one chance to convert future demand into permanent entitlement at no extra cost. The teams that plan it as a project certify the largest defensible numbers. The teams that leave it late certify whatever is visible.
Around twelve months before the certification date. Discovery, contract review, deployment of planned capacity, and evidence assembly all take time, and capacity has to be genuinely running before the window closes. Starting a year out lets each workstream run properly rather than as a scramble in the final weeks.
Five run in parallel: independent discovery of the whole estate, contract and clause review to decide what qualifies, a deployment programme to bring planned capacity inside the window, evidence assembly behind every number, and certification preparation. Each feeds the defensible count you declare at exit.
Book a confidential assessment and we will build the maximization plan that fits your estate and your certification date.